More genetic data are needed before populations are mixed: response to ‘Sourcing Eurasian beaver Castor fiber stock for reintroductions in Great Britain and Western Europe’
Halley described three options for sourcing Eurasian beavers Castor fiber for reintroduction in Britain and Western Europe: (i) use animals from populations within a single western evolutionarily significant unit (ESU); (ii) mix animals from two or three western ESUs; and (iii) mix animals from eastern and western ESUs. Option three contravenes International Union for Conservation of Nature guidelines, so debate should focus on options one and two. We believe there is a need for further genetic analysis, before a final decision can be made. This decision should not be heavily influenced by cost or be determined by the genotypes of existing captive or escaped populations.
In a review recently published in Mammal Review, Halley (2011) discussed the sourcing of Eurasian beavers Castor fiber for potential reintroduction programmes in Britain and Western Europe. Eurasian beavers have recently been the subject of a controlled reintroduction to Scotland, UK as part of a 5-year scientific trial, the outcome of which will be presented to the Scottish government who will then decide on future full-scale reintroductions. Feasibility studies have recently been published for England and Wales (Gurnell et al. 2008, Halley et al. 2009), where there is strong interest in beaver reintroduction within conservation circles. Therefore, the subject of Halley's paper is interesting, relevant and will be the focus of an important discussion for future sourcing of reintroduction stocks. The purpose of this response was not to contest the conclusions of Halley (2011). However, there are specific, key areas where there is some disagreement and potential for misinterpretation.
Here we expand on certain points and advocate further research on the consequences of intentional mixing of beaver populations in order to source reintroductions. Careful consideration of all evidence presented should receive serious attention by those involved in captive breeding and reintroduction programmes. In our opinion, a more comprehensive review should include discussion of the points we raise.
The current level of molecular genetic information available for Eurasian beavers supports the differentiation of eastern and western Eurasian lineages; however, there remain insufficient data to draw conclusions on the relative levels of genetic diversity within each extant population. The use of mitochondrial DNA data alone to examine diversity and define evolutionarily significant units (ESUs) may be misleading. The level of variation within the nuclear genome is of equal or greater importance. Studies of major histocompatibility complex (MHC) variation have shown diversity among populations, but not within populations (Ellegren et al. 1993, Babik et al. 2005), supporting a pattern of bottlenecked populations throughout Europe. However, such MHC patterns may also be influenced by local adaptation to pathogens. The only nuclear genetic data that potentially reflect truly neutral variation in the Eurasian beaver were produced nearly 20 years ago and were used to compare three populations using a total of 52 individuals (Ellegren et al. 1993). While greater nuclear genetic diversity was observed in Russia than in Norway, the overriding conclusion of this work was to warn against generalizations concerning the importance of genetic variability for the survival of an endangered population. This message is as relevant now as it was then. As stated by Halley (2011), until further DNA research is undertaken, the pragmatic approach of reintroducing unmixed stock is a precautionary recommendation. Reintroductions to date have demonstrated that successful population growth has arisen from unmixed stock [Telemark (C. f. fiber), Rhone (C. f. galliae) and Elbe (C. f. albicus) populations], therefore reintroduction failure through inbreeding depression seems unlikely (Halley 2011).
A number of initiatives are currently underway to explore the genetic diversity of the various potential source populations for UK reintroductions in much greater detail than has been previously possible. Until additional comparative data regarding the relict Norwegian, French and Elbe populations are available, any conclusions regarding the relative levels of population diversity, and their importance with respect to characterizing the fitness of the respective source populations, are premature.
INBREEDING, FECUNDITY AND GENETIC ABNORMALITIES
Halley discusses data, which strongly suggest that ‘unmixed’ beaver populations may display higher instances of inbreeding depression and phenotypic abnormalities than ‘mixed’ populations. However, it has been argued that inbreeding depression in individual fitness components is of limited importance to conservation biology, unless these reductions in individual fitness translate into reduced population growth rates (Keller et al. 2007). Also, inbreeding tolerance may be more prevalent than is currently thought (e.g. De Luca & Cocroft 2008). Moreover, attempts to mitigate possible inbreeding through population mixing may induce outbreeding depression (Kokko & Ots 2006, Edmands 2007), with the disruption of locally adapted gene complexes.
However, it is not even clear that Eurasian beaver populations have historically suffered from inbreeding depression. Other confounding factors may be responsible for some of the abnormalities cited by Halley, such as environmental pollution from endocrine-disrupting chemicals. These could contribute to abnormalities or affect fecundity in animals in Eastern Europe, and have not been investigated (e.g. Baillie et al. 2003). There are now approximately 120000 animals in Sweden descended from only 11 breeding females reintroduced from Norway, and around 70000 beavers in Norway descended from a relict population of around 120 individuals. None of the more typical indications of inbreeding (abnormalities such as cleft lips or palates, polydactyla, etc.) has been observed. Parker et al. (2002) examined 143 Eurasian beavers shot in south-east Norway that were descended, like all present-day Norwegian and Swedish beavers, from the relict population of Telemark beavers. Animals' ages were determined from patterns of tooth eruptions and root closure, or from counts of cementum annuli (van Nostrand & Stephenson 1964); jaws were therefore inspected in detail. None of these animals displayed any sign of dental or other physical abnormalities commonly associated with inbreeding depression (Howard Parker, pers. obs.).
Halley states that the work of Saveljev and Milishnikov (2002) suggests that the differences in litter size observed between ‘mixed’ and ‘unmixed’ populations are due to inbreeding. Fecundity in the North American beaver C. canadensis is affected by many factors, including e.g. food availability, elevation and female age (Novak 1987). Studies on hunted populations of North American beavers have shown that fecundity is density dependent (Novak 1987). Increased litter size is a common response to harvesting in mammals (see Tuyttens & Macdonald 2000). Embryo counts in pregnant beavers from hunted populations increase with harvest rates, up to heavy harvest rates of 70% (Hendry 1966). Litter size comparisons should therefore be made between populations with similar hunting pressure. It should also be noted that many different techniques are used to count kits and estimate beaver populations, with considerable variation in results (Novak 1987, Rosell et al. 2006). Saveljev and Milishnikov (2002) provide figures for average litter size (see their Table 1). However, the method(s) used to determine litter size are not stated, and a statistical analysis of the differences in mean values is not presented.
Many mammals use specific chemical cues in mate choice and species recognition (e.g. Roberts & Gosling 2004). Species recognition through olfactory cues acts as a behavioural isolating mechanism in some mammals (Wyatt 2003). Since subspecies represent sources of genetic variation within species, and may be intermediate stages in the process of speciation, their loss could have significant implications for species evolution and biodiversity. This raises issues of subspecies protection and biodiversity conservation priorities. Within Eurasian beavers, Rosell and Steifetten (2004) suggest that geographical isolation has led to specific chemical signalling. Telemark beavers respond for significantly longer and more strongly to castoreum from Telemark beavers than to that from Elbe beavers; differences are also supported by chemical analysis. Encouraging hybridisation while not taking into account possible geographical isolating mechanisms should be cautioned against.
ANIMAL HEALTH AND WELFARE
The reintroduction process is stressful, particularly for wild animals. Capture, handling and transport have been shown to have negative physiological consequences for animal welfare, which in turn can affect the success of reintroductions (Teixeira et al. 2007). Nolet et al. (1997) suggested that stress involved in the translocation of beavers to the Netherlands may have encouraged the onset of the disease rates experienced through weakening the animals' immune systems.
The health of source populations may have implications for the success of reintroductions and animal welfare (Mathews et al. 2006). Introduced animals can carry a range of diseases and parasites, and current evidence suggests differences between regions and subspecies. For example, Rosell et al. (2001) did not find Giardia, Cryptosporidium, Campylobacter or Salmonella in Telemark beavers (n = 133), whilst Salmonella has been recorded in beavers from Germany (Elbe beavers) and Russia (Voronezh beavers; Ramasov 1992). Infectious diseases, in particular yersiniosis and leptospirosis, were the most important cause of death (50%) in Elbe beavers translocated to the Netherlands (Nolet et al. 1997). A more thorough review of parasites and pathogens from the various populations would be welcomed. Additionally, effects of human pollutants on animal health and body condition may reveal interesting differences. For example, mean cadmium concentrations in kidneys from Elbe beavers are the greatest reported in free-ranging herbivores and about five times above the critical concentration at which kidney damage can occur in mammals and birds (Nolet et al. 1994). In contrast, the concentrations recorded in Telemark beavers were considerably below those associated with known detrimental effects in mammals (Fimreite et al. 2001). A careful consideration of both population and individual health and welfare should be made before selecting animals for reintroduction projects. Increased selectivity at the onset of reintroduction programmes will improve animal welfare in the long run.
Beaver stocks suitable for reintroduction to an island like Great Britain may not be optimal for countries in mainland Europe where populations from different stocks have the potential to mix more freely. The three options described by Halley for source stocks for reintroduction are concise and are representative of the potential scenarios if beavers are to be reintroduced to Britain. In our opinion, using a mixed eastern and western ESU source stock (option three) contravenes International Union for Conservation of Nature guidelines, and in the clear existence of alternatives, should not be implemented. The immediate debate should focus on options one and two (to reintroduce beavers from single or mixed western ESU populations, respectively). While economic practicalities should be taken into consideration, we do not believe decisions on the source stock should be heavily determined by cost or by the genotypes of captive or potential escaped populations. It is hoped that both Halley (2011) and this response will open up the discussion of provenance, and highlight the need for further data collection and analysis, before final decisions can be made at governmental levels. Until further and more conclusive genetic work is undertaken a precautionary approach is recommended.
We thank Simon Girling, Rob Ogden and Ross McEwing for valuable comments.